The overarching goals of this proposal are twofold: 1) to support the candidate in gaining additional training in the use of molecular and statistical genetic methods to identify risk alleles for substance use disorders (SUD), and 2) conduct a genome wide association study (GWAS) for a novel phenotype, a pre-morbid liability index for SUDs, and replication studies in two independent samples. The advantage of focusing on indicators of pre-morbid risk (i.e. prior to initiation of substance use) is that analyses are based on underlying rather than expressed risk and so may provide a more sensitive measure of genetic liability than manifest or post-morbid measures of SUDs. Focusing on pre-morbid liability also provides a sound theoretical framework to examine developmental processes associated with the emergence of SUDs such as exposure to environmental risk (gene-environment correlations) and sensitivity to environmental adversity (gene x environment interactions). The training portion of the proposal focuses on gaining skills in bioinformatics to identify candidate gene systems and select informative polymorphisms (e.g. SNPs) as well as expertise in statistical analyses to detect associations between genetic markers and complex phenotypes such as SUDs. The research plan entails conducting a GWAS of 61 OK SNPs on a measure of pre-morbid liability to SUDs on approximately 6000 participants. Replication studies of significant findings from the GWAS will then be conducted in two independent samples of high-risk youth. The research plan also includes studies examining mechanisms of G-E interplay that underlie the development of SUDs by utilizing a longitudinal twin design. The award will also assist the candidate in pursuing his long-term goal of becoming an independent researcher and is likely to yield important insights into the etiology of SUDs. PUBLIC HEALTH RELEVANCE: We propose a novel approach to identifying genes associated with substance abuse by focusing on risk factors present prior to initiation of substance use. Pooling multiple longitudinal samples (N~6000), we will conduct a genome wide association study on a novel measure of pre-morbid risk for substance abuse to identify specific risk genes.